YouTube Playlist Tracker

Track changes to YouTube Playlists

Features:

detect videos missing from playlist, either due to playlist owner removing the video from the playlist, or due to the video owner deleting the video from youtube.
keep backup of all playlist and video metadata
remember title and description of deleted videos

detect private videos
detect duplicate videos
notice new videos added
notify about changes to title or description of a video
update stored playlist and video metadata
flag to delete missing videos from backup

track all playlists of a channel
detect newly created playlists
detect missing playlists
flag to delete missing playlists from backup

config file for setting the channel id / list of playlist id's to query

Missing Features:

does not explicitly detect changes to playlist metadata e.g. changes to playlist title

How to use

API Authentication

This program uses the "YouTube Data API v3" by Google to get the information about channels, playlists, and videos.
To use the API, you need a Google account. You can follow these steps to get started.

If you want to track changes to any non-private playlist, or to public playlists belonging to a channel, then authentication with just an API key is enough.
If you want to track playlists that are set to private, for instance your own playlists, then you need to use OAuth 2.0 Client IDs, otherwise you won't be able to see those playlists.

API Key Files:

api_key.json

This file contains the API key you generated in the step above, to use the YouTube Data API v3. You can find this key in your google dashboard. Remember to put the key in quotes, to make it valid json.
Your file should have this format:

"this_is_my_api_key1232131"

client_secret.json

This file contains the OAuth 2.0 Authentication information. Specifically it contains the "Client ID" and the "Client Secret". Both can be found in your google dashboard, the same as with the API key above.
This file can also be downloaded from the dashboard. Just make sure to rename it, because the file you recieve from google might have a lot of junk appended to the name.
The downloaded file will have some more information, which we are not gonna use. The important information is the ID and the Secret. Your file should have this format (ignoring the useless information):

{
    "installed":
    {
        "client_id": "some-random-string-of-numbers-and-letters.apps.googleusercontent.com",
        "client_secret": "random-secret-key123"
    }
}

Config File:

tracked_ids.json

This file contains this ids of all the channels whose playlist you want to track, and a list of all the single playlist you want to track.
An example file is also included.
Please keep in mind that private playlists of a channel can only be tracked if that channel is authenticated via the OAuth 2.0 method, even if you are the owner of that channel. The API Key is not enough for that. Your file should have this format:

{
    "channels" : ["channel_id1", "channel_id2"],
    "playlists" : ["playlist_id1", "playlist_id2", "playlist_idN"]
}

Temporary File:

credentials.pkl

This is a temporary file generated when you use the save_credentials() function. It's only use is to allow you to quickly restart your kernel, without having to re-authenticate, when using OAuth 2.0.

Data File:

playlist_data.json

This file is generated when you first save playlist data. It is the file that contains the entire backup. It is technically in a human readable format, meaning you can open this file and browse it, it is however not nicely formatted, and contains a lot of superfluous information about the playlists and videos like e.g. thumbnail links. For that reason it is recommended to explore the data programmatically.
The structure of the file looks similar to this (excluding some of the information deemed superfluous):

{
    "random_playlist_id_1" : {
        "id" : "random_playlist_id_1",
        "snippet" : {
            "channelId" : "random_id_of_channel_that_owns_this_playlist",
            "title" : "The Title of this playlist",
            "description" : "The Description of this playlist",
            "channelTitle" : "Title of the channel that owns this playlist"
        },
        "videos" : {
            "random_video_id_1" : {
                "snippet" : {
                    "channelId" : "random_id_of_channel_that_owns_this_playlist",
                    "title" : "The Title of this video",
                    "description" : "The Description of this video",
                    "channelTitle" : "Title of the channel that owns this playlist",
                    "position" : 123,
                    "resourceId" : {"videoId" : "random_id_of_this_video"}
                }
            },
            "random_video_id_2" : {},
            "random_video_id_3" : {},
        }
    },
    "random_playlist_id_2" : {},
    "random_playlist_id_3" : {},
}

Usage

Take a look at the code. It's not much, and should be self explanatory. The notebook also contains lots of markup with tips, and the code is thoroughly commented.
After choosing which of the two Authentication methods you will use, setting it up, and deciding on which channels and playlists you're interested in, in the tracked_ids.json file, you should be able to run the notebook top to bottom (only execute the Authentication cells you need). The main() function is the one that does all of the heavy lifting, and the save_playlist_data() function, which is defined at the top of the notebook, can be used after main(), to store the data generated by it.

flpeters / YouTube_Playlist_Tracker